Sync conversations

Skill yinjialu/context-harness/skills/sync-conversations

Local-first personal AI context loop for Code Agent conversations

Install

npx -y skills add yinjialu/context-harness --skill sync-conversations

Assembled from the repository path, not quoted from the project. Check it against their README if it does not work.

One thing to look at

3 stars3 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.

What its author says it does

Copied from the file, not written here

Sync local Codex and Claude Code conversation records into a context-harness data home. Use when the user asks to back up Code Agent conversations, sync Codex sessions, sync Claude Code sessions, run full or incremental conversation backup, or inspect context-harness sync status.

The file declares its own license as MIT. That is the author’s claim about this one file, and it is not the same thing as the license GitHub reports for the repository, which is listed with the other numbers below.

SKILL.md

3.0 KB, as published. Nobody here has run it

Sync Conversations

Use this skill to collect local Code Agent conversations into context-harness. The skill can bootstrap the CLI runtime even when only the skill was installed.

Runtime Bootstrap

Before running context-harness, execute the bundled bootstrap script from this skill directory:

runtime_dir="$(bash scripts/bootstrap.sh)"
cd "$runtime_dir"

The script clones or updates the runtime repository at ~/.local/share/context-harness by default, checks out the latest released tag, and runs uv sync.

Overrides:

CONTEXT_HARNESS_RUNTIME_DIR: custom runtime checkout path.
CONTEXT_HARNESS_REPO_URL: custom fork URL.
CONTEXT_HARNESS_REF: custom branch, tag, or commit.
CONTEXT_HARNESS_BOOTSTRAP_SKIP_UPDATE=1: skip fetch/pull.

Sources

V1 supports:

codex
claude-code

Claude Web, Antigravity, ChatGPT, Gemini, and browser exports are outside V1.

Workflow

Run the Runtime Bootstrap steps above. Use the returned path as the context-harness repository root.
Resolve the context home from CONTEXT_HARNESS_HOME, the user's explicit path, or ~/.context-harness.
Change to the runtime repository root.
Run the requested sync command from the repository root.
Summarize checked, created, updated, skipped, and output_dir.

Commands

Before running any command, first run the Runtime Bootstrap steps and change to the returned repository root. Run all uv run context-harness ... commands from that root.

Incremental sync:

uv run context-harness --context-home <context-home> sync codex --latest 1
uv run context-harness --context-home <context-home> sync claude-code --latest 1

Full sync:

uv run context-harness --context-home <context-home> sync codex --all
uv run context-harness --context-home <context-home> sync claude-code --all

After running, summarize checked, created, updated, skipped, and output_dir.

OKF Compliance

Archives are native Open Knowledge Format files: each archive carries an OKF frontmatter block (type: Conversation, plus title/description/source/session/messages/created/tags/timestamp) above the unchanged body. Every sync also rebuilds the OKF index.md / log.md files under conversations/ and the data-home root index.md.

To upgrade a pre-OKF data home (archives without frontmatter), run the idempotent migration once:

uv run context-harness --context-home <context-home> migrate-okf --dry-run   # preview
uv run context-harness --context-home <context-home> migrate-okf             # apply

The migration also normalizes memory/*.md and global-claude.md frontmatter. It preserves unknown producer fields and human-authored bodies, so it is safe to re-run.